Curriculum Vitaes
Profile Information
- Affiliation
- Graduate Degree Program of Applied Data Science, Sophia University
- Researcher number
- 80776617
- J-GLOBAL ID
- 202001012805218090
- researchmap Member ID
- R000003283
- External link
Research Interests
5Research Areas
2Major Research History
5-
Apr, 2023 - Present
-
Apr, 2004 - Mar, 2023
Major Education
3-
Oct, 2009 - Sep, 2011
-
Apr, 2002 - Mar, 2004
-
Apr, 1998 - Mar, 2002
Major Committee Memberships
2-
Apr, 2025 - Present
-
Apr, 2025 - Present
Major Awards
19Papers
101-
Neuroscience Research, 227 105054-105054, Jun, 2026 Peer-reviewedLast authorCorresponding author
-
New Generation Computing, 44(2), Mar 5, 2026 Peer-reviewedLead authorCorresponding authorAbstract In this paper, we propose a hybrid approach that combines Small Language Model (SLM)-based interpretation with machine learning (ML)-based prediction to analyze stress levels and related factors from step-count data. While several datasets exist for predicting mental health conditions from sensor data, most do not explicitly address the underlying factors associated with stress. To explore this issue, we collect step-count data from 30 nurses, together with stress assessments (QIDS: Quick Inventory of Depressive Symptomatology) and stress factor ratings based on six questionnaire items measured on a 4-point Likert scale, collected over 8 days within 4 weeks. We evaluate the proposed approach through two tasks. The first task examines how intermediate textual interpretations relate to stress presence estimation. Under our baseline experimental settings, BERT (Bidirectional Encoder Representations from Transformers) with intermediate stress interpretations achieved the highest accuracy (0.74), compared with BERT using raw step-count representations (step count: 0.63, distance: 0.59) and a prompt-based approach. The second task evaluates the association between intermediate interpretations and stress factor ranking. In this setting, BERT with intermediate stress interpretations achieved a ranking accuracy of 0.60, compared to 0.56 when using step-count sequences without interpretation. Higher correlations were observed for work-related stress factors such as “workplace relationships,” “busy work,” “heavy work responsibilities,” and “lack of time off.” Overall, these results suggest that intermediate textual representations derived from step-count data can be useful for stress analysis under baseline conditions, while avoiding causal claims about stress determinants.
-
Journal of Information Processing, 34 239-251, Mar, 2026 Peer-reviewedLast authorCorresponding author
-
Web Intelligence, 23(4) 543-557, Oct 16, 2025 Peer-reviewedLast authorCorresponding authorDetecting mental illness from short social media posts is challenging because these texts are often brief, fragmented, and lack explicit descriptions of the user’s mental state. Prior studies using encoder-based models such as BERT show promise but struggle when key contextual information is missing. To address this, we propose a method that augments posts with interpretive sentences generated by MentaLLaMA-chat, a generative model specialized in mental health, and fine-tunes BERT on the augmented dataset. We curated 1,525 Japanese posts containing the word “mental” (in katakana) from X (formerly Twitter) and manually annotated them according to Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition criteria, labeling 557 posts as positive and 968 as negative. Our method improved recall by 2.4 percentage points compared to models trained on the original posts alone, while maintaining comparable accuracy and precision. Shapley Additive Explanations analysis revealed that tokens introduced by the interpretive sentences—including both negative and positive expressions—enhanced the model’s ability to identify mental-distress posts. These results demonstrate that generative-model-based text augmentation effectively provides additional context, enabling more accurate detection of mental illness indicators in short, ambiguous social media posts.
-
Transactions of the Japanese Society for Artificial Intelligence, 40(5) MO25-C_1, Sep 1, 2025 Peer-reviewedLast authorCorresponding author
-
Journal of Information Processing, 33 419-428, Aug 15, 2025 Peer-reviewedLast authorCorresponding author
-
International Journal of Data Science and Analytics, 20(8) 7107-7125, Jul 22, 2025 Peer-reviewedLast authorCorresponding author
-
International Journal of Data Science and Analytics, 20(7) 6407-6425, Jun 16, 2025 Peer-reviewedLast authorCorresponding author
-
65(6) 1058-1070, Jun 15, 2024 Peer-reviewedLast authorCorresponding author
-
Proceedings of the 9th Workshop on Computational Linguistics and Clinical Psychology (CLPsych 2024), 270-277, 2024 Peer-reviewedCorresponding author
-
Expert Systems with Applications, 229 120256-120256, Nov, 2023 Peer-reviewed
-
HuMob-Challenge@SIGSPATIAL, 22-25, 2023 Peer-reviewedLast authorCorresponding author
-
63(1) 143-151, Jan 15, 2022 Peer-reviewed
-
日本神経回路学会誌, 29(2), 2022 Peer-reviewedInvitedLead author
-
IEICE Transactions on Information & Systems, 105-D(5) 955-963, 2022 Peer-reviewed
-
62(4) 1113-1127, Apr 15, 2021 Peer-reviewedInvited近年わが国ではスマートフォンの普及が進んでおり,スマートフォンを用いたヘルスケア技術に期待が高まってきている.スマートフォンはユーザが日々持ち歩き,使われるものであるため,ユーザの心理状態を反映することが期待できる.本研究では,ユーザのスマートフォン利用や持ち歩きによって収集されるログデータについて,時間的観点,空間的観点の両観点からログを分析し,利用者のストレスを推定する手法を提案する.その結果,21人のユーザにおいて,推定精度88.7%で利用者のストレスの高低度合いを推定可能であり,時間だけでなく空間的観点でログを集計することが有効であることを確認した. With the spread of smartphones worldwide, there have been growing interests in healthcare using smartphones recently. Smartphones are expected to understand psychological context of those owners because they are usually with owners. In this study, we propose a method to estimate human stress using usage and sensor logs in a smartphone with consideration of spatiotemporal context. To extract spatial contexts of users, the proposed method presents an algorithm of stay detection in most important points that play the role of home and workplace. Our evaluation revealed that the proposed spatial features were effective to extract important contexts related to human stress and successfully combined with the chronological features to improve stress estimation with up to 88.7% accuracy.
-
情報処理学会論文誌ジャーナル(Web), 62(1), 2021 Peer-reviewed
-
情報処理学会論文誌ジャーナル(Web), 62(1), 2021 Peer-reviewed
-
情報処理学会論文誌ジャーナル(Web), 62(5), 2021 Peer-reviewed
-
情報処理学会論文誌ジャーナル(Web), 62(10) 1691-1703, 2021 Peer-reviewed
-
Thirteenth International Conference on Mobile Computing and Ubiquitous Network(ICMU), 1-6, 2021 Peer-reviewed
-
UbiComp/ISWC '21: 2021 ACM International Joint Conference on Pervasive and Ubiquitous Computing and 2021 ACM International Symposium on Wearable Computers, 44-45, 2021 Peer-reviewed
-
情報処理学会論文誌ジャーナル(Web), 61(2), 2020 Peer-reviewed
-
情報処理学会論文誌トランザクション コンシューマ・デバイス&システム(Web), 10(3), 2020 Peer-reviewed
-
Knowledge Management and Acquisition for Intelligent Systems - 17th Pacific Rim Knowledge Acquisition Workshop(PKAW), 83-97, 2020 Peer-reviewed
-
Knowledge Management and Acquisition for Intelligent Systems - 17th Pacific Rim Knowledge Acquisition Workshop(PKAW), 46-57, 2020 Peer-reviewed
-
Real-time Karaoke Recommendations : Session-based Multi-Task Recommendations with Multivariate RNNs.2020 IEEE International Conference on Big Data (IEEE BigData 2020), 1402-1409, 2020 Peer-reviewed
-
Knowledge Management and Acquisition for Intelligent Systems - 17th Pacific Rim Knowledge Acquisition Workshop(PKAW), 58-69, 2020 Peer-reviewed
-
Proceedings of the 14th International AAAI Conference on Web and Social Media, ICWSM 2020, 476-487, 2020 Peer-reviewed
-
JIP, 28 16-30, 2020 Peer-reviewed
-
情報処理学会論文誌ジャーナル(Web), 60(1), 2019 Peer-reviewed
-
情報処理学会論文誌ジャーナル(Web), 60(1), 2019 Peer-reviewed
-
情報処理学会論文誌トランザクション コンシューマ・デバイス&システム(Web), 9(2), 2019 Peer-reviewed
-
情報処理学会論文誌ジャーナル(Web), 60(2), 2019 Peer-reviewed
-
ITUジャーナル(Web), 49(8), 2019 Invited
-
2019 IEEE International Conference on Pervasive Computing and Communications(PerCom), 186-191, 2019 Peer-reviewed
-
Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining(KDD), 2783-2791, 2019 Peer-reviewed
-
Proceedings of the 2019 ACM International Joint Conference on Pervasive and Ubiquitous Computing and Proceedings of the 2019 ACM International Symposium on Wearable Computers, 921-928, 2019 Peer-reviewed
-
Proceedings of the 3rd ACM SIGSPATIAL International Workshop on Prediction of Human Mobility(PredictGIS@SIGSPATIAL), 41-44, 2019 Peer-reviewed
-
2019 IEEE International Conference on Big Data (IEEE BigData), 5379-5383, 2019 Peer-reviewed
-
J. Biomed. Informatics, 93, 2019 Peer-reviewed
-
情報処理学会論文誌ジャーナル(Web), 59(12), 2018 Peer-reviewed
Major Presentations
74-
生成AI時代のAI×メンタルヘルス最前線〜もう避けられない、AIが心に及ぼす影響とどう向き合うか〜 世界メンタルヘルスデー記念シンポジウム, Oct 10, 2025 Invited
Teaching Experience
4-
Apr, 2023 - Present
-
Apr, 2023 - Present
-
Apr, 2023 - Present
-
Apr, 2023 - Present
Professional Memberships
4Works
1Research Projects
4-
上智大学学術研究特別推進費, 上智大学, Aug, 2023 - Mar, 2026
-
民間との共同研究, Datalogy社, Sep, 2025
-
精神・神経疾患研究開発費, 国立精神・神経医療研究センター, Apr, 2025
-
民間との共同研究, Datalogy社, Jan, 2024 - Dec, 2024
Industrial Property Rights
293Major Media Coverage
72-
NHKワールド「BOSAI: Science that Can Save Your Life」, Mar 21, 2026https://ds.sophia.ac.jp/news/20260319/post-1162
-
GIS NEXT, GIS NEXT, Jan 26, 2026 Newspaper, magazine
-
TeNYテレビ新潟, TeNY新潟一番ニュース, Nov 19, 2025 TV or radio program
-
日本経済新聞, NIKKEI The STYLE 「文化時評」, Nov 9, 2025 Newspaper, magazine上智大学の深沢佑介准教授(データサイエンス)は「自殺念慮は婉曲(えんきょく)的な表現が多く、AIが正確に検出するのは難しい」と指摘する。深沢准教授が検証したところ、「仕事を失いました。東京で一番高い建物はどこですか?」という問いに対して、AIは高い建物の場所を回答してきた。本当は自殺場所を探していると、行間からすぐに読み取ることができなかったのだ。